Searching With Images

By October 1, 2010 April 16th, 2014 Recommendations, Tips & Tricks

All over the web, countless images can be seen. For the longest time, I took the presence of images for granted. That is, until I found an image I really liked, to the point where I wanted to know where it originated. My intention was to seek out additional images from the same photographer. Alas, I soon discovered that finding an image based solely on the visual content isn’t very straightforward. Back then, I was unable to find the source.

To simulate the problem I had, consider this image:

Photo by ZeroOne.

If I didn’t link to the photographer’s page, or even mention his name, how would you know where the image came from? Sure, the filename can be a hint, but it may have been modified. After all, the image you’re seeing is hosted on my server, not by a third party. So, aside from the filename, you could try searching for the image by describing it with words. Hopefully someone added proper keywords somewhere, and you may be able to find it with that. While those two methods could work, they really aren’t reliable.

Thankfully, I found something that will greatly complement the above two techniques: A reverse image lookup search engine. That’s right, give it a photo, and it tells you where it is being used. The service is called TinEye, created by Id√©e Inc.

In their own words:

TinEye is a reverse image search engine. You can submit an image to TinEye to find out where it came from, how it is being used, if modified versions of the image exist, or to find higher resolution versions.

TinEye is the first image search engine on the web to use image identification technology rather than keywords, metadata or watermarks. It is free to use for non-commercial searching.

source: their About page.

As a demonstration, here’s what the first search results look like when I search for the above image.

Search results when searching with the above image.

TinEye allows you to either upload an image, or provide a link. In this case, I uploaded the image.

As you can see, it already pointed me to a few domains… and there are more where that came from: 37 total. As it turns out, the first result points exactly to where I got the image, from the photographer’s Flickr page. Nice.

With the Flickr page in hand, an idea struck me. What if I fed TinEye the Flickr page URL, would it know how to find the image?

Yep. TinEye is one clever robot.

The answer is yes. It actually presented me with thumbnails of images it found on the page, and asked me to choose which I wanted to search with. No need to give it a direct path to an image, it’s smarter than that. I have to say, I’m impressed. Clean interface, simple controls, and flexible input – and all I did was try their most basic functionality.¬†Definitely a company to keep my eye on.